A short history of the internet, in five turning points

The internet and the web are often used as if they mean the same thing. They do not: the internet is the global network, and the web is one service that runs on it. Their histories run side by side.

Before 1969: the idea of packets

In the 1960s, computers were large, expensive machines, and researchers wanted ways to share them. Telephone networks worked by holding open a dedicated circuit for each call, which was wasteful for computer traffic that came in short bursts.

A different idea emerged, developed independently by Paul Baran in the United States and Donald Davies in Britain: packet switching. Instead of reserving a line, a message is broken into small packets, each labelled with its destination. The packets can take different routes and are put back together at the other end. If one route fails, packets can travel another way. That idea remains at the heart of the internet.

1969: ARPANET’s first message

ARPANET, funded by the US Defense Department’s research agency, connected its first computers in 1969. The first message, sent from the University of California, Los Angeles to the Stanford Research Institute, was supposed to be “LOGIN”. The system crashed after “LO”.

By the end of that year, four sites were connected. Within a few years, dozens more had joined, and the network’s users found an application its designers had not planned around: in 1971, Ray Tomlinson sent messages between computers on the network, choosing the @ sign to separate a user’s name from their machine. Email quickly became one of the network’s most popular uses.

1983: one common language

On 1 January 1983, ARPANET switched over to TCP/IP, the set of rules that lets different networks pass data between them. That common language is what made an internet of networks possible, and it still underpins it today.

TCP/IP was designed by Vint Cerf and Bob Kahn, who published their approach in 1974. Its key insight was that the network itself should be simple, passing packets along without needing to understand them, while the computers at each end take responsibility for checking that data arrives correctly. That design made it easy for new kinds of networks, and later new kinds of applications, to join.

The Domain Name System followed in the 1980s, letting people use names such as example.org instead of numeric addresses.

1989: the World Wide Web is proposed

At CERN, the European particle physics laboratory, Tim Berners-Lee wrote a proposal for an information system using hypertext. By the end of 1990 he had built the first web server and browser. CERN put the web software in the public domain in 1993, which helped it spread.

The web rested on three simple ideas that are still in use: addresses for each document, now called URLs; a way to request documents, called HTTP; and a format for writing them, called HTML, in which links can point to any other document on any other server. Nobody needed permission to add a page or link to someone else’s.

The 1990s: browsers and the public

Graphical browsers made the web easy to use, and commercial internet access spread to homes. Search engines arrived to help people find their way around the rapidly growing number of sites.

Mosaic, released in 1993, showed text and images together on the same page and was simple enough for non-specialists. Commercial browsers followed. At the same time, the backbone networks of the internet moved from government funding to commercial providers, opening the door to online shops, banks and news sites. The rapid rise in technology company valuations late in the decade, and the crash that followed in 2000 and 2001, became known as the dot-com bubble.

The 2000s onward: always connected

Broadband, then smartphones, turned the internet from something you sat down to use into something you carry. Today most people reach the web on a phone, which is why sites, including this one, are designed for small screens first.

The launch of the iPhone in 2007, followed by Android phones, put a full web browser and app store in millions of pockets. Social networks, video streaming and cloud services grew on top of that foundation, and much of daily life, from banking to booking a doctor’s appointment, moved online.

Common misconceptions

• The internet was not one invention. It was built step by step by many people in universities, government labs and companies over several decades.
• ARPANET was not mainly a nuclear-war shelter. Survivable communications motivated some early research on packet switching, but ARPANET itself was largely built to let researchers share computing resources.
• Nobody owns the internet. It is a network of networks run by many organisations, coordinated through open standards and bodies such as the Internet Engineering Task Force.

What is still uncertain

The internet’s next turning points are harder to name in advance. Billions of people still have no regular access to it, and how quickly that gap closes is uncertain. Questions about privacy, about the power of a few very large platforms, and about how countries regulate online content remain under active debate. The open design that let the internet grow so quickly is also what makes those questions so difficult to settle.